Search

Visual illustration transformation from real-world to cartoon images is one of the famous and challenging tasks in computer vision. Imageto-image translation from real-world to cartoon domains poses issues such as a lack of paired training samples, lack of good image translation, low feature extraction from the previous domain images, and lack of high-quality image translation from the traditional generator algorithms. To solve the abovementioned issues, paired independent model, high-quality dataset, Bayesianbased feature extractor, and an improved generator must be proposed. In this study, we propose a high-quality dataset to reduce the effect of paired training samples on the model’s performance. We use a Bayesian Very Deep Convolutional Network (VGG)-based feature extractor to improve the performance of the standard feature extractor because Bayesian inference regularizes weights well. The generator from the Cartoon Generative Adversarial Network (GAN) is modified by introducing a depthwise convolution layer and channel attention mechanism to improve the performance of the original generator. We have used the Fréchet inception distance (FID) score and user preference score to evaluate the performance of the model. The FID scores obtained for the generated cartoon and real-world images are 107 and 76 for the TCC style, and 137 and 57 for the Hayao style, respectively. User preference score is also calculated to evaluate the quality of generated images and our proposed model acquired a high preference score compared to other models. We achieved stunning results in producing high-quality cartoon images, demonstrating the proposed model’s effectiveness in transferring style between authentic images and cartoon images. [ABSTRACT FROM AUTHOR]

Ideally, Convolutional Neural Networks (CNNs) should be trained with high quality images with minimum noise and correct ground truth labels. Nonetheless, in many real-world scenarios, such high quality is very hard to obtain, and datasets may be affected by any sort of image degradation and mislabelling issues. This negatively impacts the performance of standard CNNs, both during the training and the inference phase. To address this issue we propose Wise2WipedNet (W2WNet), a new two-module Convolutional Neural Network, where a Wise module exploits Bayesian inference to identify and discard spurious images during the training and a Wiped module takes care of the final classification, while broadcasting information on the prediction confidence at inference time. The goodness of our solution is demonstrated on a number of public benchmarks addressing different image classification tasks, as well as on a real-world case study on histological image analysis. Overall, our experiments demonstrate that W2WNet is able to identify image degradation and mislabelling issues both at training and at inference time, with positive impact on the final classification accuracy. • A novel two-module CNN exploiting Bayesian inference is introduced. • We address image degradation and mislabelling both at training and at inference. • Our method overcomes the deterministic counterpart (standard DenseNet121). • We show results on two public benchmarks and a real-world histological case study. [ABSTRACT FROM AUTHOR]

Bayesian computation plays an important role in modern machine learning and statistics to reason about uncertainty. A key computational challenge in Bayesian inference is to develop efficient techniques to approximate, or draw samples from posterior distributions. Stein variational gradient decent (SVGD) has been shown to be a powerful approximate inference algorithm for this issue. However, the vanilla SVGD requires calculating the gradient of the target density and cannot be applied when the gradient is unavailable or too expensive to evaluate. In this paper we explore one way to address this challenge by the construction of a local surrogate for the target distribution in which the gradient can be obtained in a much more computationally feasible manner. More specifically, we approximate the forward model using a deep neural network (DNN) which is trained on a carefully chosen training set, which also determines the quality of the surrogate. To this end, we propose a general adaptation procedure to refine the local approximation online without destroying the convergence of the resulting SVGD. This significantly reduces the computational cost of SVGD and leads to a suite of algorithms that are straightforward to implement. The new algorithm is illustrated on a set of challenging Bayesian inverse problems, and numerical experiments demonstrate a clear improvement in performance and applicability of standard SVGD. [ABSTRACT FROM AUTHOR]